Tuning of Multivariable Model Predictive Control for Industrial Tasks

This work is concerned with the tuning of parameters Model Predictive Control (MPC) algorithms when used for industrial tasks, i.e., compensation disturbances that affect process (process uncontrolled inputs and measurement noises). The discussed simulation optimisation procedure quite computationally simple since consecutive are optimised separately, it requires only a very limited number simulations. It makes possible to perform multicriteria control assessment as few quality measures may be taken into account. effectiveness method demonstrated multivariable distillation column. Two cases considered: perfect model case more practical in which characterised by some error. shown approach obtain good quality, much better than most common all constant.